Batch file extracting certain tokens from unknown number of delimiters in variable - batch-file

Question

Third, and hopefully final revamp of question...
Batch file reading text file with for loop line by line into variable. Each line of said text file can be formatted completely different from the next. The only common delimiter would be a four-digit number (year) somewhere in each line. The goal is to return whatever text follows the aforementioned four-digit number for each line via echo.
Text file example:
Monday, January 1, 1900 there was an event-e6718
On this day in 1904 nothing occurred
Wednesday, March 3, 1908 an error occurred when attempting to access the log
Thursday, , 1911 - access denied
Friday, in whatever month, on whatever day, in 1938, nothing happened
Therefore, based on the above text file example, the return would be like...
there was an event-e6718
nothing occurred
an error occurred when attempting to access the log
- access denied
nothing happened
As of 1318 PST, I've tried every code snippet in the comments following, but none of them were able to return the data I need to return.
However, those comments were related to my initial question which has since been markedly improved.
I've even tried a regular expression of "^[1-9][0-9][0-9][0-9]$", however I'm new to regex, so I'm sure I have that wrong.
Is this even possible?
Thanks in advance.

Answer 1

Batch is a horrible task for this. REGEX is a good tool, but cmd doesn't support it (with the exception of a very crippled subset with findstr). If you are willing to use an external tool, it gets easy:
<old.txt call jrepl ".*(\d{4})\D\ *(.*$)" "$2" >new.txt
search for a four digit number \d{4}, followed by a non-digit \D and zero or more spaces, and take the rest until "EndOfLine" .*$. (Parentheses) mark matches, referenced by $x. Your desired strings are in $2.
Output with your example file:
there was an event-e6718
nothing occurred
an error occurred when attempting to access the log
- access denied
there was an event-dsfd318
nothing happened
Should you decide to include the year, you can find it in $1:
<old.txt call jrepl ".*(\d{4})\D\ *(.*$)" "$1: $2" >new.txt
gives:
1900: there was an event-e6718
1904: nothing occurred
1908: an error occurred when attempting to access the log
1911: - access denied
1910: there was an event-dsfd318
1938: nothing happened
The call is neccessary wihtin a batch file, because jrepl is a batch file and therefore wouldn't return without call.
(The REGEX pattern might be subject for improvement; I don't have much experience with it yet.)
jrepl.bat was programmed by dbenham.

Answer 2

Give this a go:
#echo off & setlocal enabledelayedexpansion
for /f "delims=" %%i in ('type "C:\textfile.txt" ^| findstr /IRC:"there was an event"') do (
set "event=%%i"
echo "!event:*there was an event=there was an event!"
)
textfile.txt
Monday, January 1, 1900 there was an event-e6718
On this day in 1904 nothing occurred
Wednesday, March 3, 1908 an error occurred when attempting to access the log
Thursday, , 1911 - access denied
Monday, January 1, 1910 there was an event-dsfd318
Friday, in whatever month, on whatever day, in 1938, nothing happened
Result:

Answer 3

If there truly is no common point in the string or consistency in token count, adjust the number of iterations in the below for loop to match the maximum possible tokens.
#Echo off & Setlocal EnableDelayedexpansion
Set "event=Monday, January 1, 1900 there was an event-e6718"
For /L %%i in (1 1 10) Do (
Set "event=!event:*, =!"
)& rem // arbitrary number of iterations that should be adjusted to match the maximum expected tokens
Set "event=%event:~5,100%"& rem // remove the year[space] from the string - final string maximum length is also arbitrary and may need adjusting.
Echo/%event%
** Update **
An example of a macro to get the last token within a for loop using a macro version of the above approach:
Note: you'll need to adjust the file path for yout input file.
#Echo off
(Set \n=^^^
%=Newline Var=%
)
Set Gettoken=For %%n in (1 2) Do if %%n==2 (%\n%
For /F "Delims=" %%G in ("!string!") Do (%\n%
Set "output=%%G"%\n%
For %%D in ("!Delim!") Do For /L %%i in (1 1 10) Do Set "output=!output:*%%~D=!"%\n%
Set "output=!output:~5,100!"%\n%
)%\n%
Set output%\n%
) Else Set string=
Setlocal EnableDelayedExpansion
Set "Delim=, "&& For /F "Delims=" %%I in (inputfile.txt) Do %GetToken%%%I
Edit:
single batch solution for the amended questions actual requirement.
#Echo off & CD "%~dp0"
Setlocal Enabledelayedexpansion
rem // replace inputfile.txt with your filepath
For /F "Delims=" %%L in (inputfile.txt) Do (
Call :sub "%%L"
rem // the below for loop will remove everything up to including the first year from the string
rem // as well as traling coma[space] / [space]
For %%E in (!Errorlevel!) Do (
If Not "%%E"=="0" (
Set "String=!String:*%%E=####!"
Set "String=!String:####, =!"
Set "String=!String:#### =!"
Set "String=!String:####=!"
)
)
rem // output only if a year "delimiter" was encountered
If not "%%~L"=="!String!" Echo/!String!
)
Exit /b
:sub
Set "String=%~1"
rem // adjust for loop %%I for valid year range and %%# for maximum expected string length
For /L %%I in (1899 1 2050) Do (For /L %%# in (0 1 100) Do (If "!String:~%%#,4!"=="%%I" (Exit /B %%I)))
Exit /B 0